data1701d

I think you give valid examples and make your point well.

However, another weird thought is perhaps we’re always slowly dying to some extent. For instance, you at age 7 is dead; today, yourself at age 7 cannot speak or act or think. For instance, in a situation where your young self may have tried to buy a toy, you have different wants and make different decisions - you cannot perfectly replicate what that past self would have wanted.

This might be true even of myself from five seconds ago - I hadn’t thought of a certain wording of this concept yet, and so might have worded it differently under different circumstances - that “me” is gone and can’t do anything. This could be true even a millisecond ago, or a duration approaching either an instant or perhaps one cycle based on whatever the “clock rate” (if there is such a thing) or the human brain is.

However, to function, we need a convenient abstraction for what life and death are. I think my definition of life would be when one particular sum of experiences permanently terminates its (mostly) granular evolution.

Thomas and Will Riker both evolved from the same sum of experiences of the original William T Riker; since those sums of experience are still evolving, he is, within our convenient definition, alive.

Our family actually has a bunch that an aunt sent once.

Cool. In a little over a month, I hit 3 years.

Was about to cite TNG Tech Manual as well - although that also said that holodeck characters’ bodies were replicated meat puppets, which I think they didn’t stick with.

If it doesn’t simulate a connected monitor, it looks like there are little HDMI shims that do called EDID emulators that are available for relatively cheap.

(Note: Anything I say could be B.S. I could be completely misunderstanding this.)

Clevis isn’t too difficult to set up - Arch Wiki documents the process really well. I’ve found it works better with dracut that mkinitcpio.

As for PCR registers (which I haven’t set up yet but should), what I can tell, it sets the hash of the boot partition and UEFI settings in the TPM PCR register so it can check for tampering on the unencrypted boot partition and refuse to give the decryption keys if it does. That way, someone can’t doctor your boot partition and say, put the keys on a flash drive - I think they’d have to totally lobotomize your machine’s hardware to do it, which only someone who has both stolen your device and has the means/budget to do that would do.

You do need to make sure these registers are updated every kernel update, or else you’ll have to manually enter the LUKS password the next boot and update it then. I’m wondering if there’s a hook I can set up where every time the boot partition is updated, it updates PCR registers.

I got a response to my e-mail; they say it’ll be fixed by Monday.

JavaScript be like that sometimes…

That is so me sometimes.

You're somewhat right in the sense that the point of disk encryption is not to protect from remote attackers. However, physical access is a bigger problem in some cases (mostly laptops). I don't do it on my desktop because I neither want to reinstall nor do I think someone who randomly breaks in is going to put in the effort to lug it away to their vehicle.

Clevis pretty much does TPM encryption and is in most distros' repos. I use it on my Thinkpad. It would be nice if it had a GUI to set it up; more distros should have this as a default option.

You do have to have an unencrypted boot partition, but the issues with this can at least in be mitigated with PCR registers, which I need to set up.

It’s a smidge more difficult on Debian if you want to use a non-ext4 filesystem - granted for most people, ext4’s probably still fine. I use it on my desktop, which doesn’t have encryption.